2 stroke thread (with occasional F1 relevance!)

[manolis]

Hello J.A.W.

You write:
"
Actually, in the view of David Vizard, the role of the exhaust in a high output 4T application - is of major importance..
He writes:
"Compared with intake, exhaust tuning is far more potent, & can operate over 10 times as wide an rpm band."
"

The role of the exhaust "in a high output 4T application" is of major importance.



but not in normal green 4-stroke engines, wherein the extended overlap is a problem, not a solution.


I think here is wherein the point is missed:

Nobody can argue that the 2-strokes can make lots of power.
However they are problematic as regards their emissions.
The focus must be on the emissions.
On anything that can lower the emissions, even at the expense of power.


The PatATE suggests a substantially different way of operation of the 2-strokes.
And combines several unique characteristics not met in other 2-strokes (like the cold exhaust (which means, among others, substantially lower specific lube consumption), like the extreme rate of exhaust area opening, like the substantially earlier closing of the exhaust etc)


Since the lawn mower engine



is the favourite reference here, let's see how it handles the charge.

The exhaust port opens early, otherwise the crankcase, when the piston opens the transfer port, will fill with burnt gas.
After the end of the transfer, the still wide open exhaust port provides a "free way" to the fresh mixture to escape. Say as the Atkinson - Miller cycle does in the 4-strokes, but here anything that leaves the cylinder is lost and pollutes.
At the end of the exhaust, a good part of the entered air or air-fuel-mixture is not in the cylinder, and the torque drops dramatically.


And here is how things work in the PatATE:

The exhaust can open later; the rate the exhaust opens is substantially higher than in a conventional allowing the faster blow-down even with substantially later exhaust opening.
From the beginning the PatATE has a gain (on thermal efficiency and on torque) due to the extended expansion allowed by the later exhaust opening.

At the end of the blow-down the transfer opens and the scavenging starts.
The air / air-fuel mixture from the crankcase gathers progressively speed and inertia (as happens during the overlap of the sport/racing 4-strokes, article of David Vizard), with the pressure in the crankcase dropping progressively.
The entering air-mixture expels the burnt gas out of the cylinder.

Then the exhaust closes. The exhaust can close more than 50 crank degrees earlier than in the conventional lawn mower 2-stroke engine.

During the closing of the exhaust, the rotary valve of the PatATE opens the intake port (without a reed valve to delay the begining of the intake stream).

The vacuum created in the crankcase due to the inertia of the air / mixture going from the crankcase to the cylinder, suctions new air /mixture into the crankcase through the open intake port. Like a supercharger, but for free.

When the transfer port finally closes, the quantity of air /mixture trapped into the cylinder for compression / combustion / expansion has nothing to do with a lawn mower.

According the previous analysis, the quantity of air /mixture that can be trapped into the cylinder of the 2-stroke PatATE (or, differently, the volumetric efficiency) is comparable with the quantity of air /mixture trapped in the cylinder of the sport-racing 4-strokes, like the Ducati Panigale.
And if you double the specific torque of the Panigale (because the PatATE is a 2-stroke), the expected specific torque becomes more than 30% higher than in the green KTM EXC250TPI / 2018.


Where is the mistake or the trick?

Does the PatATE design fit with lawn mower engines or with green super-2-strokes?

Thanks
Manolis Pattakos

[Pinger]

The vacuum created in the crankcase due to the inertia of the air / mixture going from the crankcase to the cylinder, suctions new air /mixture into the crankcase through the open intake port. Like a supercharger, but for free.

Just look at the pumping efficiencies of the two competing chambers, one with a pressure ratio of 7.5:1 (minimum), the other with circa 1.4:1 and you believe that inertia will overcome the cylinder's ability to reverse the transfer flow and direct it back to the crankcase? Not a chance.

According the previous analysis, the quantity of air /mixture that can be trapped into the cylinder of the 2-stroke PatATE (or, differently, the volumetric efficiency) is comparable with the quantity of air /mixture trapped in the cylinder of the sport-racing 4-strokes, like the Ducati Panigale.

I think you need to understand the difference between 'perfect displacement' (unachievable) scavenging and 'diffusion' in scavenging. When you do, you will realise that 4T levels of volumetric efficiency are impossible in a 2T (without copious (costly) over supply of purging air) and not required as the firing on every stroke coupled to other merits (eg, low friction, bulk, mass, etc) are compensation enough.

If you want 4T volumetric efficiency - build a 4T.

[J.A.W.]

Hi Manolis..

Yes, you are correct to note volumetric efficiency in sports/race performance engine 'powerband' parameter terms..
..so the Ducati Panigale uses its vast engine capacity/light weight - to overcome low rpm torque deficiencies..

But variable valve timing/flow control.. has been available for decades.. in both 2T & 4T road-vehicle usage..
& Uniflow has demonstrated (in-metal/empirically in operation) the efficacy of 2T TFI & variable rotary inlet valve tech..

[manolis]

Hello all.

I read:

“I think you need to understand . . .. When you do . . . ”

also:

"build one and show us"



To show that an engine is euro4 or euro5 compliant is far more difficult and costly than measuring the fuel consumption.

And it took almost 20 years from the filing of the 2-stroke TPI patent of Rotax (previous post) till a 2-stroke TPI to be officially tested on the roads by KTM (reasonably, KTM has to pay royalties to Rotax).



Let’s try, without making it in metal, to understand how things work in the 4-stroke Ducati Panigale, and then apply the results in the PatATE.

Here are the torque curves of the Panigale 1299 and 1199 (the dashed line is for the 1199):



At 9,000rpm the 1199 makes 110mN/lit specific torque and keeps it till 10,500rpm of the peak power.

At only 4,000tpm it makes “just” 90mN/lit specific torque, which is good for most 4-strokes.

The “hole” from 4,500 to 7,000rpm seems deep only because the peak torque is so high.


Worth to note:

There is neither VVA (Variable Valve Actuation) nor VVT (phaser) in this top technology engine.
Just four big valves per cylinder actuated desmodromically.


Quote from http://www.cycleworld.com/2014/09/12/te ... ght#page-9



“9) ...while the 1199 Superleggera uses titanium for intake and exhaust also by Del West. Each titanium exhaust valve saves 24 grams over its steel counterpart on the 1199 Panigale. Reduced inertia means the engine is capable of more revs. Valve lift is unchanged: 16mm intake, 14.3mm exhaust. There is a slight difference in the intake cam timing, although duration remains the same. The Superleggera intake cam is rotated 4 degrees backward, opening 15 degrees BTDC and closing 64 degrees ABDC (19 and 60 degrees, respectively, on the Paginale unit).”


In the following plot (Piston Speed versus Crank Angle):



the piston speed at 15 degrees after the TDC is less than 1/3 of the maximum piston speed at middle stroke.

No matter what, the inertia of the exhaust helps substantially the filling of the cylinder.

During the overlap, the inertia of the gas in the exhaust creates a vacuum that suctions, through the opening intake valves, air /mixture accelerating it.
When the exhaust valves finally close (15deg ATDC), the air / mixture coming from the intake finds the exhaust closed and a still slow moving piston.
The inertia of the “intake column” increases the pressure just before the intake valves and inside the cylinder (the kinetic energy of the “intake column” turns to “dynamic energy”).
And as the piston accelerates towards the BDC the pressure drops and the speed of the air / mixture from the intake valves increases.


Despitethe fact that the exhaust and the intake were “met” for only 30 degrees (and with a substantially low average valve lift during the 30 degrees of the overlap), the job was done by turning the velocity to pressure and back to velocity.

The Panigale Superleggera makes 200hp from two cylinders and 1200cc.



Now, please read again my last post and write where the mistake / trick is.

After the paragraph:

“When the transfer port finally closes, the quantity of air /mixture trapped into the cylinder for compression / combustion / expansion has nothing to do with a lawn mower.”

It can be added:

“Just before the closing of the transfer port, and due to the inertia of the entering air / mixture, the pressure in the cylinder is higher than in the crankcase which, at the same time, is fed, with air /mixture entering through the intake port.
After the closing of the transfer, the flow through the intake port to the crankcase continuous (due to the inertia, again) becoming stronger and stronger as the piston moves towards the TDC (due to the crankcase sub-pressure, this time).
The filling of the crankcase continues for several (say 60?) degrees after the TDC, even after the crankcase pressure is above the ambient pressure, due to the inertia again.
Then the intake port closes by the rotary valve of the PatATE and the compression in the crankcase starts.”

Thanks
Manolis Pattakos

[Tommy Cookers]

don't mistake me for a person who has said anything against Manolis's 2 stroke

regarding the Panigale and Superleggera plots ......

is the given valve timing correct ? (ie is it run timing not eg values at a special large valve clearance for timing definition only ?)
because the declared overlap of 30 deg or 38 deg seems far too small for such a very high performance engine spec

nearly 60 years ago BSA made a similar niche version (single cylinder) road machine - it had 115 deg of valve overlap
unsurprisingly it could not reach maximum speed unless taken to peak power rpm through the gears
the Ducati torque plots show the same fault (that would occur in a single cylinder version)
they are bizarrely unsuited to road-normal use (like the RD350 as exposed here, or worse)
afaik they would not be allowed in F1 torque mapping

hasn't the position been that claims in Vizard's article apply only with a race or similar high-overlap cam ?

[gruntguru]

Tommy if you look closely at the torque plots provided you will see that the Panigale 1299 has between 108 and 144 Nm across the entire rev band shown (4k -11k rpm). That is not a "peaky" engine.

[Tommy Cookers]

the torque falls from 4000 to 5500 rpm - this fall is surely an indisputable manifestation of 'peakiness' ?

for road driving we need a torque curve consistently rising or flat (at least not falling) over most of the speed range
because only this will allow a consistent response to the right foot action (or inaction) that we call driving
eg where hills and/or headwinds appear, the right foot demands an instant and seamless torque increase

we might by borrowing an aviation analogy see this consistent torque curve as essential to 'speed stability'
the traditional rising torque curve at the low speeds and falling torque curve at the higher speeds giving easier and better control of speed
speed stability is artificially enhanced eg in airliners by automatic thrust control on landing approaches, as is apparent on windy days

even a CVT would without intelligence be less than satisfactory dealing with an inconsistent torque curve
though the undesirable driven pulley over-clamping in the simple CVT helps to stabilise its behaviour and so satisfy the user
improving the simple CVT is not easy for this reason

[Pinger]

is the given valve timing correct ? (ie is it run timing not eg values at a special large valve clearance for timing definition only ?)
because the declared overlap of 30 deg or 38 deg seems far too small for such a very high performance engine spec

Laboratory figures TC. On-road figures will be higher. That's VAG ownership for you....

[Pinger]

for road driving we need a torque curve consistently rising or flat (at least not falling) over most of the speed range
because only this will allow a consistent response to the right foot action (or inaction) that we call driving
eg where hills and/or headwinds appear, the right foot demands an instant and seamless torque increase

we might by borrowing an aviation analogy see this consistent torque curve as essential to 'speed stability'
the traditional rising torque curve at the low speeds and falling torque curve at the higher speeds giving easier and better control of speed

The quest for fuel efficiency removes any desire to run beyond the speed of Tmax (providing the 'torque back-up' that used to make for 'flexibility') with 4Ts. Thus the flexibility has to be with the transmission's keenness to drop ratios in response to demands from a driver's right foot.
As I've alluded to before, the 2T's lack of sensitivity to throttling losses can provide what you seek without a fuel consumption penalty.

even a CVT would without intelligence be less than satisfactory dealing with an inconsistent torque curve
though the undesirable driven pulley over-clamping in the simple CVT helps to stabilise its behaviour and so satisfy the user
improving the simple CVT is not easy for this reason

I'm not convinced passive CVTs do actually suffer from driven pulley over-clamping, that particular issue being put to bed with the advent of torque sensitive driven pulleys.
What is obvious though is that the efficiency peaks at the 1:1 ratio - where the keen eye will observe that the radius of each sheave (as the belt sees it) is identical. It is moving to lower or higher ratios and incurring a smaller radius on one sheave and thus forcing more bending on the belt that drops the efficiency. Until the evidence appears to support over clamping as the the cause, I'll run with the above as the true cause of inefficiency in the higher ratios.

[uniflow]

I give up, he who shouts loudest wins, even if it was black, manolis would argue it was white, just for arguments sake.
KTM are NOT using a Rotax patent, just have a look at the patent yourself, if you can't see then perhaps new glasses might be in order. KTM are using this Kiwi TPI system because they DON'T have to pay anyone for the patent, they found it free off the internet, saw how well it worked then used it, good on them.
The second cylinder that was developed by KTM WAS the rotax patent, first being the Orbital system, this the third development is a copy of my YZ system.
For those that may not believe me, not long after we shot the youtube video of the YZ running around my yard (that would be in real running hardware), Kiwi Rider magazine did an article on this TPI system in both the F9 Kawasaki and the YZ. The guy that wrote the article used to work at KTM Austria and sent Kiwi Rider magazines back to his friends at KTM for coffee table reading. This Kiwi TPI development was handed to them, the KTM development team were able to read about it at lunch time ( I explained the system fully in the article), finally for KTM a workable answer AND they didn't have to pay anyone royalties, a win for KTM I say.
But keep on believing what you want.

[gruntguru]

I'm not convinced passive CVTs do actually suffer from driven pulley over-clamping, that particular issue being put to bed with the advent of torque sensitive driven pulleys.
What is obvious though is that the efficiency peaks at the 1:1 ratio - where the keen eye will observe that the radius of each sheave (as the belt sees it) is identical. It is moving to lower or higher ratios and incurring a smaller radius on one sheave and thus forcing more bending on the belt that drops the efficiency. Until the evidence appears to support over clamping as the the cause, I'll run with the above as the true cause of inefficiency in the higher ratios.

I think Manolis' claims of over-clamping as a cause of friction relate to the "cheap" moped CVT market.

[manolis]

Hello Uniflow.

Several pages ago you had asked what differs the OPRE / PatOP from the other Opposed Piston engines.

This is exactly what a patent is for:

A patent is an official document wherein they are defined the differences of a new idea from the prior art, and the advantages these differences brink to the art.


The patent examiners have the authority and the duty to search, examine and decide whether the idea in question is really new and patentable.
Nobody else has such authority.
If required, the patent examiners ask the inventor /applicant to limit his invention in order what is left, to be really his intellectual property.

This is how things work.

You can’t claim intellectual property over an idea / invention in the US without applying from a patent in the US-PTO and without a patent to be eventually granted for the specific idea / invention.

Only after the granting of a patent the inventor can claim that this, or that, is his own idea / property.


And it costs way less than making a good quality working prototype.


To put it in numbers:

Here are the fees paid so far for the PatEf:



In a few months the UK_IPO will respond by the list of the most relevant patents they will find with their search.

Then, we will decide to proceed with the substantive examination, too (the additional cost of 80 UK pounds makes the total cost for taking the UK patent: 230 UK pounds, or US300$), or to withdraw the patent.


Are the 150 pounds too much?

Shouldn’t you apply for a patent before making your TPI engine prototype?

What can you do now to support that the TPI is yours and not of KTM or of Rotax?


On the other hand, applying for a patent is not so simple.
It means that you can define the differences of your idea from the hundreds of the TPI patents mentioned by the examiner in the TPI patent granted to Rotax some 17 years ago (previous post).


So, what are the differences of your TPI?

What makes it unique among the other TPI’s proposed so far?

To put it differently:
If Rotax asks you to pay royalties (or asks you to stop using your prototype), have you strong arguments to refuse?


As I wrote again, this is the way things work in the “intellectual property” field.

Even if you don’t like the previous "teaching", it may prove useful and profitable (in the long term) for some guy who is silently reading this forum and is having a project in mind.

The advice is simple and general: "apply for a patent first, and then reveal your idea".



By the way:
I know that I know nothing about engines.
This is why I asked the others, who think they know, to spot on the mistakes / tricks of my simpleminded approach about the PatATE operation based on the inertia of the working gas.
I am still waiting.



PS.
What would be the estimated cost for manufacturing a PatATE prototype?
E-mail me.

Thanks
Manolis Pattakos

[manolis]

Hello Pinger.

You write:
“What is obvious though is that the efficiency peaks at the 1:1 ratio - where the keen eye will observe that the radius of each sheave (as the belt sees it) is identical. It is moving to lower or higher ratios and incurring a smaller radius on one sheave and thus forcing more bending on the belt that drops the efficiency. Until the evidence appears to support over clamping as the the cause, I'll run with the above as the true cause of inefficiency in the higher ratios.”



This plot:



was made based on the data provided in the publication:
http://www.pattakon.com/PatBox/SECVT.pdf
of the:
Eindhoven University of Technology
Department Mechanical Engineering
Dynamics and Control Technology Group

The above plot concerns the most expensive (and technologically advanced) CVT for scooters: the SECVT of the Suzuki Burgman 650 and of the Aprilia Mana 850 (made/used under the license of Suzuki).

As you can see, at the medium transmission ratios (red curve, 1:1 ratio: the radiuses the V-belt runs on the two pulleys are, more or less, equal) the efficiency is not higher than at the lower transmission ratios (cyan curve) wherein happens “more bending on the belt that drops the efficiency.”

The efficiency is about the same (red curve and cyan curve) at the heavier loads, while at the lighter loads the CVT is substantially more efficient at the lower transmission ratios.

In the longer transmission ratios (blue curve) the CVT operates inefficiently.


Let’s put it in numbers to see what “inefficiently” means:

The Suzuki Burgman 650 makes 55hp at 7,000rpm, and 62mN at 5,000rpm (which means, at 5,000rpm it makes 44hp).

For “highway cruising” (say 100Km/h, lever road, no opposite wind) it needs no more than 10hp of power (a medium size car needs around 15hp), and the transmission ratio cannot help being near the longest possible.

According the above plot, the efficiency of the SE-CVT of Burgman at cruising is somewhere between 80% and 85% (far below the 95% peak efficiency), which means the scooter consumes some 10-15% more fuel.
For the 10hp the CVT passes to the rear wheel at cruising, another 1.5hp is consumed inside the CVT as friction overheating the V-belt compartment and wearing the V-belt.


Now, take another look at the PatEf.

The PatEf simply unloads the V-belt CVT (of all scooters and sleds, expensive or not) from a significant part of the unnecessary clamping (and friction, and wear it causes), without any risk of belt slippage.



As it is now, the most expensive scooter CVT (the SECVT) operates at cruising not at the FLA point (above plot), but with three times heavier clamping of the V-belt (top right point).

Thanks
Manolis Pattakos

[Tommy Cookers]

what I called driven pulley over-clamping seems to degrade transmission % (power?) efficiency by the effect of excessive belt tension ......
when the drive pulley is at a large diameter so there's a lot of belt length in contact

but the power loss/belt heating is no greater at these low efficiencies than at the high efficiency end (because that's high power)

reduced clamping will best benefit fuel consumption only where in cruise the engine is unthrottled or slightly throttled
so might be considered an alternative to a conventional design of greater ratio range and bulk ?

to self - I notice that modern car CVTs are rigged to give some rpm rise with roadspeed (presumably to match buyer prejudice)

[manolis]

Hello Tommy Cookers

You write:
“but the power loss/belt heating is no greater at these low efficiencies than at the high efficiency end (because that's high power)”


You are right.

And this means that:

If with 20kW power provided by the drive shaft, the CVT “eats” the 1kW in friction / wear and delivers the rest 19kW to the driven shaft,

and if with 5kW power provided by the drive shaft, the CVT “eats” again 1kW in friction / wear and delivers the rest 4kW to the driven shaft,

then

in the first case the efficiency is 95%,

while In the second case the efficiency drops to 80%.




From another viewpoint:


Take the (coloured) plot showing the SECVT efficiency versus the load.

With the PatEf the over-clamping drops substantially at the longer transmission ratios, it drops at the medium transmission ratios, and it remains the same at the lower transmission ratios,

So, the cyan curve (efficiency at the lower transmission ratios) remains unchanged.

But the other two curves change.

According the theory behind the PatEf:

the red curve (efficiency at the medium transmission ratios) will shift upwards, above the cyan curve for all loads, remaining above, say, 96% from 60% load to 100% load,

while the blue curve (efficiency at the longer transmission ratios) will shift substantially upwards, at or above the cyan curve for all loads, remaining above 95% along the heavier load range,

say, like:



If the above are correct, the CVT (of scooters, sleds, bicycles etc) “plays” as equal with the conventional gear-boxes as regards the efficiency, being undoubtedly superior as regards the easy handling / user friendly.

Thanks
Manolis Pattakos